Neural Network Reactor Control for Degradation and Transient Response
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Nuclear power plant control systems face challenges in consistently achieving desired responses due to component degradation and transient events, leading to safety margins and suboptimal operation.
Innovation Solution
A neural network is trained using a simulated nuclear power plant environment to determine optimal control settings, accounting for various plant parameters and conditions, enabling precise control of steam generator levels and other critical parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional control systems are used to control nuclear reactor components, then the system structure is simple and easy to understand, but the control precision deteriorates due to component degradation and transient events
Solution Approach 1:
A neural network is introduced as an intermediary component between the control system and the reactor components. The neural network receives sensor data as input and outputs optimized control settings, acting as a mediator that bridges the gap between traditional control systems and the complex, degraded plant conditions. This intermediary layer enables precise control despite component degradation without requiring complete system redesign.
Solution Approach 2:
The patent replaces traditional mechanical/control system approaches with an intelligent software-based neural network. Instead of relying on fixed control algorithms and mechanical adjustments, the system uses a trained neural network that can adaptively determine optimal control settings based on learned patterns from simulation data, substituting rigid mechanical control with flexible intelligent control.
2Reliability
If safety margins are increased to ensure safe operation during component degradation, then safety is improved, but productivity deteriorates due to operating below maximum power
Solution Approach 1:
The neural network enables dynamic control settings that adapt to current plant conditions rather than relying on fixed conservative safety margins. By continuously learning from simulation data and adjusting control recommendations based on real-time sensor inputs, the system can dynamically optimize the balance between safety and productivity, allowing operation closer to maximum power when conditions permit while maintaining safety.
Solution Approach 2:
The system changes control parameters based on learned patterns from simulation data. The neural network analyzes various plant states and determines optimal control parameter adjustments that maintain safety while maximizing power output. This parameter optimization allows the plant to operate more efficiently without compromising safety margins.
3Reliability
If conservative control settings are used to account for component degradation, then reliability is improved, but manufacturing precision deteriorates in achieving desired plant response
Solution Approach 1:
The neural network is trained in advance using extensive simulation data that covers a wide range of plant conditions, including component degradation scenarios. This preliminary training action allows the network to learn optimal control strategies before actual plant operation, enabling it to provide precise control recommendations that account for degradation without requiring conservative trial-and-error adjustments during real-time operation.
Solution Approach 2:
The system implements a feedback mechanism where the neural network receives actual sensor data from the plant and adjusts its control recommendations accordingly. This closed-loop feedback allows the system to maintain precise plant response by continuously comparing actual conditions with learned patterns and making real-time control adjustments, overcoming the limitations of static conservative settings.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A method of training a neural network to determine at least one control system setting for a nuclear power plant (10) comprises providing a simulation (102) of the nuclear power plant; generating training and test data by inputting a plurality of control system settings and nuclear power plant parameters into the simulation of the nuclear power plant and obtaining simulated sensor data and a simulated plant response from the simulation of the nuclear power plant; and generating a trained neural network (200) trained to determine the at least one control system setting for the nuclear power plant to achieve a desired plant response for the nuclear power plant using the training and test data.